clock menu more-arrow no yes

Filed under:

UConn men’s basketball using new data analysis tool to gain upper hand on opponents

The Huskies are one of a growing number of D-I programs utilizing ShotQuality’s analytics to improve their competitive advantage.

NCAA Basketball: Big East Conference Tournament-Creighton vs Connecticut Vincent Carchietta-USA TODAY Sports

There is a data analysis tool, ShotQuality, being used by eight NBA teams and over 40 Division I college basketball programs that was created by a 20-year-old college student.

At its core, ShotQuality is a program that automatically analyzes the quality of every shot of every college basketball game in the country. It logs each shot and gives it a score using over 90 different variables, which can then be viewed by the teams that purchase it.

Basketball is a game of skill, but there is quite a bit of luck involved as well. This tool was made to quantify how much luck is involved and find out what the likely outcomes of each shot and game are.

“To put a game into perspective is really important for a coach,” CEO and Founder Simon Gerszberg said. “They need to understand the fact if this was a lucky outcome or if we just took terrible shots and didn’t deserve to win.”

Gerszberg started off calculating the quality of shots by hand for Colgate’s men’s basketball team, where he works as a data analyst, using a formula originally created by Colgate assistant coach Dave Klatsky. He was able to automate it for the program at the end of last season then did so for all of men’s college basketball over quarantine.

He launched the tool in October 2020 and has already attracted the attention of basketball programs like Stanford, Texas, Xavier, and, more recently, UConn.

Part of what has made the product attractive to these programs so far is how it presents the data. There are in-depth numbers on the website, but it also describes the data collected in more understandable terms – like shot charts and lists of the best shots in each game.

There is a statistic for good and bad possessions, for example. A good possession is one that results in a shot in the 75th percentile or better while a bad possession is one in the 25th percentile or below.

“I think what he’s done, especially on the college level, is really broken it down and taken a lot of the analytics and put it in layman’s terms, for lack of a better term,” UConn’s Director of Recruiting and Video Tripp Doherty said. “He’s made it a lot more understandable.”

UConn’s coaching staff felt the team was getting solid looks at the basket, but noticed the Huskies had a lower close two-point percentage than expected, Doherty said.

Doherty reached out to Gerszberg over Twitter and bought a ShotQuality subscription for the program on Feb. 23, according to the purchase agreement obtained from the university under the Freedom of Information Act. Since then, UConn has gone 5-1 with the Huskies’ only loss coming against Creighton, a fellow ShotQuality customer, in the Big East Tournament semifinals.

ShotQuality has been most useful for scouting UConn’s opponents before games, Doherty said. The scouting staff looks at each upcoming opponent’s tendencies – what shots they take and where – and then relays that information to the players to use on the court.

“I think the individual personnel player scouts percentile ranks has helped me in our scouting reports for opponents so much,” Doherty said.

UConn makes defensive game plans each game, but much of what they do is based on head coach Dan Hurley’s coaching principles. They use the ShotQuality website to have information ready for pregame and in-game tweaks if they need to shift something, Doherty said.

“The more information that we can have going into a game, the better we feel about it,” said Doherty. “In turn, the players feel better about it because they know that they’re going in with all the information they could possibly have to be successful.”

There are two subscription models available for purchase on the ShotQuality website right now. One is for coaches of college basketball and NBA teams and the other is a new feature, which gives consumers access to a predictive model that gives game-by-game score predictions.

“There’s multiple teams in college basketball that are actually being overvalued and undervalued by Vegas, because of three-point lock mostly, but other stuff too,” Gerszberg says.

This feature is split into two pages: the predictions and March Madness tabs. The predictions page breaks down each scheduled game on every given day. The March Madness tab looks at every possible match-up in this year’s tournament and presents data for how that game could play out.

ShotQuality has a betting Twitter account as well, where they make daily betting picks for men’s college basketball games using this predictive model. At the time of writing, their model is 177-154-1 against the spread.